A Compressed Domain Dual Video Watermarking for Real-Time Applications

Abstract

Main objective of dual video watermarking is to achieve better compression rates, robustness against attacks and high security which has been rarely reported. In this paper, visible and invisible watermark is embedded in the video encoder during compression, resulting in an optimized compression/watermarking algorithm and system. Due to its lower complexity, Discrete Cosine Transform (DCT) watermarking is chosen here. Exhaustive simulations made on recovered watermark had high correlation after different attacks. This algorithm is suitable for in real-time digital video broadcasting and emerging applications such as video fingerprinting.